THE EARTH IN METEORIC SHADOW. 149 



evidence in Gilbert White's weather records, in the 

 Natural History of Selbome. In 1771 he tells us of 

 spring weather and rain to the end of the first fortnight 

 in May ; in 1772, the first fortnight of May was dry, 

 with cold, piercing winds; in 1773, throughout May 

 and ^Tune, 'warm showers'; in 1774, no marked 

 peculiarities; in 1775, cold weather throughout May; 

 in 1776, cold weather throughout May; in 1778, a 

 rainy May; warm Mays thence till 1782, when the 

 first week of the month was dark and cold; in 1783 

 there was thick ice on May 5 ; in 1784, cold, dry 

 weather during the first twelve days of May ; in 1785, 

 mild weather during the first twelve days of May ; in 

 1786, mild weather during the first seventeen days; 

 on May 1 and 2, 1786, 'thick ice'; in 1787, fine 

 bright weather to the 22nd; in 1788, a warm, dry 

 May; in 1789, a warm, wet May; and lastly, while 

 May in 1790 and 1791 was warm throughout, May in 

 1792 was cold and bleak. It will be seen presently 

 that these variations in the weather of May are as 

 important a part of the evidence as the average de- 

 pression of temperature on the * cold days. 



With reference to the cold days of February, the 

 evidence is even more remarkable, as well for wideness 

 of distribution in space, as for the length of time 

 during which the phenomenon has been noticed. M. 

 St. Claire Deville, searching meteorological records for 

 evidence respecting the cold week in February (which 

 he had found to be noted throughout Europe and in 

 America), actually found that it had been noted in 



